One long-term objective is to understand the mechanisms involved in diseases, particularly glaucoma, that damages the ganglion cell/optic nerve. The multifocal VEP is a relatively new, objective, and noninvasive, technique for assessing local ganglion cell/optic nerve damage. Aim 1 seeks a better understanding of the mechanisms of glaucomatous damage by comparing multifocal VEP recordings, behavioral measures of visual sensitivity and structural measures of ganglion cell loss. To aid in these comparisons, two theoretical frameworks/models are proposed, one relates the mfVEP amplitude to behavioral sensitivity, while the other relates each of these measures to ganglion cell loss. Experiments are proposed to test these models. In addition, as it is important from the perspective of neuro-protection to know if there are ganglion cells that are functioning abnormally, evidence of "sick" ganglion cells will be sorted by examining the latency of mfVEP responses. Multiple sclerosis associated with optic neuritis also affects ganglion cell/optic nerve function. The multifocal VEP is uniquely qualified for examining local changes in conduction latencies over time. To improve our understanding of structural and functional damage due to optic neuritis, Aim 2 asks about the long-term recovery in latency that is reported to occur in some patients. To improve the utility of the multifocal VEP for both clinical and basic science, we need a better understanding of the relationship between the multifocal VEP and the underlying cortical physiology. In Aim 3, a model is proposed that relates the amplitude of the multifocal VEP to V1 cortical physiology. The assumptions of this model are tested in a series of experiments. A second long-term objective is to improve our methods for detecting ganglion cell/optic nerve damage. The experiments under Aim 4 are designed to improve on the currently available methods of recording and analyzing the multifocal VEP. The aim is to improve the efficacy of the multifocal VEP technique in the clinic. [unreadable] [unreadable]